Name Matching Is a Must for Background Screening

Name Matching, Risk Management

Name matching for background screening

To Minimize Risk of Damage, Organizations Need to Perform Background Screening

A lot is at stake for organizations who are about to hire an employee or add a new contractor or partner. They may be exposing themselves to substantial risk of legal trouble or reputational damage. A prospective employee may have a criminal background and could pose an insider threat danger. A new contractor or partner organization may likewise have past legal or other reputational issues that could damage an organization’s public image or pose real legal problems.

Background Screening in the Public and Non-Profit Sectors

This is true not only of companies but also organizations in the public and non-profit sectors. These organizations regularly perform background screening and due diligence on employees, contractors, and partners against various screening lists such as criminal records and sex-offender lists.

For example, Government intelligence and military agencies need to thoroughly monitor current and prospective employees to ensure they don’t pose a risk to sensitive or classified information.

Organizations that serve vulnerable populations (e.g., minors, patients) such as schools and hospitals need to monitor current and prospective employees as well as volunteers. In the case of schools, they need to ensure that teachers and volunteers can be trusted with the care of minors. In the case of medical organizations, they need to ensure that their staff will behave ethically in all aspects of medical care.

In the non-profit sector, charity organizations need to vet any foreign organizations they work with to ensure that those foreign organizations aren’t in reality pursuing unacceptable political goals. It’s not uncommon for organizations in conflict zones working ostensibly towards humanitarian ends to also have goals that are incompatible with those of the charity.

Background Screening in the Private Sector

Companies regularly perform background screening of employees, as well as prospective vendors and partners, against criminal records. For many companies a complicating factor is the increasing globalization of business: not only do many businesses sell across the world, but their employees, vendors, and partners can come from anywhere. Getting a clear picture of them is increasingly difficult.

Moreover, companies in many industries have compliance obligations and are impacted by laws and regulations such as Know Your Customer (KYC) and FCPA (Foreign Corrupt Practices Act). It’s critical that they establish standards and procedures for minimizing the risk that may come in the door with corrupt employees, partners, or even volunteers. There may be significant legal consequences, risk of litigation, and hefty fines if they fail to do this. Companies avert these dangers by performing background screening and due diligence on all prospective and current employees and third parties. This needs to be done against various screening lists such as sanctions lists, Anti-Money Laundering (AML) lists, and PEP (Politically Exposed Persons) lists.

Organizations need to ensure that their background screening and due diligence is as comprehensive and accurate as possible. It must uncover all the individuals and organizations likely to pose risks, but also make sure that no one is tarred unjustly. In order to do this, they need a technology that can accurately match a person’s and company’s name against the various screening lists. It also needs to be able to match other data elements pertaining to a person, such as date and place of birth, address, etc.

How Name Matching Protects a Company against Risk

Name Matching (aka Entity Resolution or Entity Matching) is an AI technology that provides a strong basis for performing background screening and due diligence. Using it, a company can screen all prospective hires, contractors, etc., against all the relevant screening lists.

Name Matching applies fuzzy matching to accomplish its goals. It uses sophisticated AI and Machine Learning technology to calculate the similarity between one data record containing multiple fields of data and another.

Each field exhibits different characteristics that have to be handled:

  • Personal names can differ in many ways:
    • Word order variations: First Name/Last Name vs. Last Name/First Name
    • Initials instead of full names: John Powell vs. J. Powell
    • Names of different ethnicities can exhibit different patterns of variation:
      • Ahmad al-Jallad vs. Ahmad Jallad (al- is a name element that can be dropped);
      • For more examples of types of variations, see here.
  • Organization names also show characteristic variations:
    • Extensive use of abbreviations: Naval Research Laboratory vs. NRL
    • The name may be shortened: Rayview Enterprises vs. Rayview
    • Presence vs. absence of corporate designators: Dixon & Howe LLC Dixon & Howe
  • Dates have their own peculiarities:
    • Different orders of elements: September 10, 2014 vs. 10 September, 2014
    • Presence vs. absence of day name: Monday, Sept. 7 1945 vs. Sept. 7, 1945
    • Full day name vs. abbreviation: Monday vs. Mon.

The above is only a sample of the variations that effective Name Matching has to handle. Other data types that it covers include the following:

  • Addresses
  • Social Security numbers
  • Place of birth
  • Account numbers
  • Credit card numbers
  • etc.

Name Matching is a Critical Technology to Support Background Screening

A state-of-the-art Name Matching product must exhibit the following characteristics:

  • Handles the types of Name Matching challenges described above
  • Provides extremely accurate matching through the use of Machine Learning and AI techniques
  • Returns a list of matches ranked in terms of similarity scores. This allows a user to set thresholds on how exact a match has to be and possibly have a human review any match in a gray zone.
  • Handles Big Data, meaning that it can query against a very large database of names in real time to support critical business applications
  • Allows the matching behavior to be tunable. For instance, a user can assign different weights on different fields based on how important the field is to the overall match.

In sum, state-of-the-art Name Matching supports effective Background Screening and Due Diligence. It will allow an organization to protect itself against bad actors.